Examine patients with osteoarthritis (OA) of the knee, as defined by ACR clinical criteria, with the needle arthroscope for typical cartilage changes of OA (fibrillation, fissures, or erosions/ulcerations) and the existence, location and intensity of the synovitis using a standard system for macroscopic evidence of synvitis. Injection of the knee with a corticosteriod after which a clinical assessment is performed by a physician who is blinded to the arthroscopy results. Correlate the degree of macroscopic synovitis at the time of arthroscopy with the clinicl response to the intraarticular corticosteroid injection. We are currently performing 3-4 arthroscopies per month and are recruiting patients into the study which should maintain this rate of accrual into the study. We are confident that we will have enrolled 30-40 patient subjects at the end of the study period in order to complete this phase of the investigations. This study consulted with the GCRC biostatistician. A modification of this study requires core lab sample processing.